Engineering Experience, Tips and Tricks for Managing a Team, and Monitoring With PMM - Percona Podcast 42

by Steve Hoffman, Matt Yonkovit

Link to listen and subscribe: PodBean

Join Percona’s HOSS (Head of Open Source Strategy) Matt Yonkovit as he sits down with Steve Hoffman, VP Engineering at Percona, to talk about his Engineering experience and his journey to Percona. Steve dives into several interesting topics including: tips and tricks for managing a team, challenges face in the open source space, the growth and development of Percona Monitoring and Management (PMM), finding and hiring engineers, and how to contribute to an open source project.

YouTube

Link: https://youtu.be/474iAPyUGy0

Steve Hoffman

VP Engineering at Percona

Technology leader who thrives in challenging, fast-paced environments committed to delivering solutions and value to both the internal and external customer. Rich mix of business management and technology coupled with demonstrated ability to lead multiple development teams on multiple platforms focusing on meeting performance, operational and financial objectives.

See all talks by Steve Hoffman »

Matt Yonkovit

The HOSS, Percona

Matt is currently working as the Head of Open Source Strategy (HOSS) for Percona, a leader in open source database software and services. He has over 15 years of experience in the open source industry including over 10 years of executive-level experience leading open source teams. Matt’s experience merges the technical and business aspects of the open source database experience with both a passion for hands on development and management and the leadership of building strong teams. During his time he has created or managed business units responsible for service delivery ( consulting, support, and managed services ), customer success, product management, marketing, and operations. He currently leads efforts around Percona’s OSPO, community, and developer relations efforts. He hosts the HOSS talks FOSS podcast, writes regularly, and shares his MySQL and PostgreSQL knowledge as often as possible.

See all talks by Matt Yonkovit »

Transcript

Matt Yokovit: Hello, everyone, welcome to another HOSS Talks FOSS, I am here with Steve Hoffman, VP of Engineering here at Percona. I’m also here with potentially the long guys who if you hear lawnmowers outside, it’s because they have decided to curse me once again with podcast visitors outside. But hopefully, they will disappear shortly. So how are you, Steve?

Steve Hoffman:
I’m doing spectacular, Matt, how are you today?

Matt Yokovit: I am good. I am good. So many of you may have not met Steve before, but he’s been here for what about two years now? Two years, two full years at Percona. And Steve runs engineering teams. And so his job is to make sure that all the software that you guys are trying out and using is stable and performant, that it has all the features that you want. So when you have something that you need Steve is the guy who helps get the team to prioritize it, Steve’s the guy who makes your life wonderful, and hell at the same time. So congratulations, Steve.

Steve Hoffman: I think I think if you ask the team, they would tell you, Steve only does the part that makes your life hell, they do the stuff that makes your life wonderful. But if you want to give me credit for making lives wonderful, I will take it.

Matt Yokovit: Alright, fair enough. So, Steve, you’ve been doing engineering for quite some time. And Percona is a little different than your background. Right? So you’ve done a lot of SaaS apps, you’ve done a lot of different applications. But coming to Percona. And jumping into this deep open-source experience, deep like remote work across the globe contributors. I mean, maybe talk to us a little bit about the difference that you’ve experienced that you’ve come here like, like, from a classic engineering job that you’ve had in the past to this new kind of thing. What was that like?

Steve Hoffman: Well, I mean, the shock for me was, one, I’ve never shipped software to a user, right, we’ve always delivered it via the web. And so the constraints that I wasn’t prepared for like, number one, we don’t have to have installation instructions. And it seems so, so simple in hindsight, but like, your mind just glosses over so many points, because you’ve done the install half a dozen times on your own. And so it’s just oh, just step one, step two, step three. And when you go to write it down, it’s sort of that classic having kids document making peanut butter and jelly sandwich where they forget to say, take off the lid and stick the knife with the butter spreading end in first, like, all of those things that we take for granted, we miss. The other side is like, you have to put twice as much effort into being correct on the first try. As I learned the hard way, once we ship it, it’s really hard to unring that bell. So above some incomplete package, a case that we hadn’t considered becomes kind of a ticking time bomb, and we don’t have the ability to go user by user and sort of force them to upgrade, right? It’s sort of how can we put that information in front of them as quickly as possible. So you know, we spent a lot of time covering the different mediums, we do it within the apps themselves as best we can, but then on our forums, and blog posts, just any way to announce hey, there’s a new version, then it fixes something we missed on the last pass. Oops. And then probably the last big one for me is the idea of breaking changes. You know, we open source I’ve never created open-source software been a big part of consuming open source certainly like to give back in the open-source. But like this, we’re actually putting it out there. And releasing open-source software, people find new and creative ways to use it, which we don’t always plan for. But that creates sort of an obligation to keep it as consistent as we can. So you know, where I’m told on a couple of instances, well, when in the SaaS world, we did it this way. So we should just convert everything over that way like Well, yeah, but it’s twice as much work because we still have to maintain the old one, like why I will, because people have built it into their automation to their continuous integration environments. Like, we can’t just change this unless you’re ready to increment to another major version like, oh, okay, so So those were some of the differences like those took a little bit of getting used to before I finally came to terms and stop fighting the fact that there are two different models for software delivery, each has their pluses, and each has their we’ll call them little surprises.

Matt Yokovit: Now we’ve got a very diverse engineering staff across the world. And that’s always a challenge to manage how to have you kind of what, what are some tips and tricks you can give to others who are starting this because now everyone’s kind of gone remote because of COVID. And you know, now, whether they go by to the office or not, maybe not. And so I’m sure that there are a lot of people out there who are looking for that kind of expertise, especially in the engineering area where it’s like, I used to have guys who all sat in the bullpen we all work together, we got together with a whiteboard, we drew out what we wanted. Now we don’t?

Steve Hoffman: Yeah, well, I mean, tip number one, if you’re not funny, in your native language, you’re even less funny in 34 different languages. So anybody who’s thinking about managing a globally diverse team takes that one to heart.

Matt Yokovit: I’m glad I’m funny in 68 different languages.

Steve Hoffman: Well, I’m still working on the first one, but I can’t get there someday. Now I know. That’s why I appreciate your mentorship. It’s helped me tremendously. But you know, exactly what you said with so many different cultures represented. There are just subtle little differences. You know, I expressions that are common, especially in the US like I’ve used this one on one on one occasion, I actually spent more time explaining globally, what a Monday morning quarterback is, then people just getting what it’s, I mean, obviously, the whole team knows it now, because I have explained so many times. But those little idioms, I guess, that we use, don’t always translate. And sometimes we just shoot right over them.

Matt Yokovit:
So you hit a home run, right? hit a home run with this one hit a home run, oh, yeah.

Steve Hoffman: And I don’t get most of the soccer reference

Matt Yokovit:
Swinging a miss, yeah, all of those.

Steve Hoffman: I’d like you said, I do miss elements of not having everybody in the room, I think the thing I liked about energy, well put it this way. The last couple of offices I’ve worked at fell in love with these open office floor plans, which I’m going to suppress my opinion on them, but one of the benefits that they have is sort of everybody in the same area. And if two dads were talking over here, and you were talking with somebody over here, you can kind of hear him say, hang on, hang on. I don’t know, I think we’re talking about something similar, right? Everybody’s swinging their chairs to the middle, quick powwow move right past whatever issue we’re facing at the time. That’s, that’s almost impossible right now, like and with the all-digital meetings, we can’t talk over each other. So there can really only be one conversation going on at a time, we are busier than ever, I don’t know how every other company is, but we’re calling it we’re busier than ever. And so if it’s very easy to get sort of the meeting fatigue, and well this really was not really about me. So I’ll just focus on the stuff that I need to get done. And I’ll keep one year open, which basically is saying I’m not listening at all. So I mean, some of those challenges, I think are significant, right. And the other side, which we learned probably as a result of the pandemic is Percona had has done a great job at really making sure that while we’re a fully remote workforce, we still make the personal connections, we still, we still have opportunity to get together face to face, have a drink, tell some jokes, have some laughs watch people go by. And the pandemic has heard that, and I don’t think we realize just how vital that is. In the grand scheme of things that it’s such a significant part of relationship building and delivering software, because for one, all the teams that Percona engineering, have a sort of a creator and consumer relationship, right the product, the PMM product, consumes our database software, our operators consume the database software, we work with our backup software. So all of these things have these interrelationships. So we have to have those connections because it’s easy to stop. And so I did my part, I did task A, B, and C, I don’t know what happens to it. Next, we’ll just ship it on down the line, and someone else will deal with it. So those are just some of those challenges. And none of them are unsolvable. But I think they take just a little bit of extra time and energy to think about what the best solutions are and sort of not stopping there and keep an open mind to things change. Maybe there are better solutions.

Matt Yokovit:
Yeah. And I mean I know, Steve, you have had many great experiences remotely and in the office. But there’s one experience that I have to ping you about about the office right now. Now. Now, the other week, Steve and I went to lunch, it was working on his 15th anniversary, and he was shocked that I knew this story. But it seems that Steve was also not only in charge of engineering, but he was also in charge of physical security at the office because he was the biggest guy at the office. Steve is a hockey player and a hockey coach. So Steve You know, just to kind of set up this story, Steve’s running engineering, he’s got that, I call it a data center, he calls it a closet, but it’s a room, it’s a room. And one day he comes in to find what Steve

Steve Hoffman: Squatter.

Matt Yokovit: So for those of you who aren’t familiar with a squatter, Steve like comes into the office, they’re noticing things missing over a few week periods, the different things that are odd right breaks

Steve Hoffman: Not quite. Not quite, I’ll back up so so we’re like, early 2000s very small startup running some office space. And our I believe at this point, we had finally stepped up to like a real data center for our production stuff, but like all of our non-production stuff, and even some of our backup stuff all ran out of the office, we had a, I think we had a T three installed at the time, and so we had decent bandwidth back and forth between the primary data centre, any rate, we took the front closet of the office, and we put an old medical equipment like an imaging server rack, and we just put all of our nonproduction stuff in our backup site stuff in there. And we call it a day you know, these are old compact proliant servers, they generated a tonne of heat, they were allowed as all get out. So we put them as far away from anybody as we possibly could, we didn’t have a receptionist. So it was just sort of convenient. We had good monitoring, right? So I wasn’t actually running engineering at the time, I was like the lead sysadmin the IT manager for the group, and I got, it was getting pages, and I couldn’t figure out what the pages were at night because it was controlled by you know, a landlord or whatever the building management company to heat shut off at night or lower to a certain temperature the light shut out so that they would conserve electricity would leave it on all night. And it was relatively cold outside but I’m getting alerts that said that the server room is overheating or our network closet for lack of a better term is overheating. And so you know I live 25 minutes away so I tried my damnedest to avoid it as long as possible and to make a long story short I went into the office and you know Bajin got through the front door and I look and I can see the ventilation system that I had sort of installed with my jigsaw and a Home Depot great was blocked on the inside with cardboard and I didn’t make any sense to me and so soon as I opened the door there was a person in there living in their living in the office and went into that room for warmth and we didn’t keep it locked for a small company and you know everybody knew it was in there was nothing exciting so yep ran down the hallway like a scared little person and screamed out like I’m going to call the police if you don’t get out and unbeknownst to us they had sort of been hiding out there for a while maybe not weeks or months but I went back and you know some rudimentary footage that we had you could see they just sort of wandered in one day. Again, unmonitored door during the day and found a place to camp out we just didn’t go in there often enough to know until the overheat page came so yeah that was my scariest moment probably ever you know

Matt Yokovit: You’re missing some of these important pieces here now you had to bribe this person to leave.

Steve Hoffman: Well I mean there was a certain threatening with the police which we thought was the more enticing one there was a bribe of food you know will wasn’t that.

Matt Yokovit: What finally sealed the deal is you had to give her your sandwich.

Steve Hoffman: it was well no I had to buy it I’d actually go get it we had a there’s a Bojangles up the road that was open till like you know 2 am or something like that which is the only thing that was open in the park at that time and so actually Okay, I’ll go out this far and so got the person into the to the lobby of the office like so they wouldn’t leave the building but at least leave the lobby so I could lock the door and it was sort of on my honor and I went and got sandwich came back and you know got a bag of I to guess it was Bojangles biscuits I don’t remember exactly what I got but I know that’s how we were able to clear the premises, no police were called and that’s how it went down.

Matt Yokovit: And now you know like I mean this is pre a lot of the physical security issues that people really you know, changed a lot of their policies and everything else but you know that’s one of those stories that you’re like yeah, that’s kind of why you have to have those watch who badges in after you and you try to opt-in after you.

Steve Hoffman: We implemented all kinds of security policies. After that, I had just a very basic way cam that was taking a picture every minute. And we call it a day. And I learned really quick to use a couple of open-source tools to string that together into a video and play it basically, I could watch the entire overnight in about three minutes every morning so that I can see Did anything happen overnight that we should be worried about. And as the software improved, we were able to do basic motion detection, which I did end up shutting off because every time a car would pull into the roundabout, their headlights would shine through. And it was just enough to change because there was no heat-sensing or anything like that was just enough change to trigger it. And so you know, again, the whole purpose was for me to sleep easy at night not get up every 15 minutes. So yeah, we had better badging, we had we put card access on the server room, or network closet door, whichever you want to call it. video cameras and a lot more places. Yeah, definitely changed the way that we treat security in that small company, we had just sort of taken it for granted up to that point, it didn’t. I suppose that the follow on is it only helped to a certain extent, right, we put all of our protections around the off-hours, and we didn’t really do anything differently during the day, and the doors auto unlocked at 8 am. So I’m more than one, it was one specific occasion, I think it was two times we had a art vendor with sort of a portfolio of artwork, who would just be lying to the back of the building. And I the one day I remember seeing her kind of blow by my door because I was toward the front but not in the front. And of course, she goes right back to like the CEO or the CTOs office or something like that is trying to sell art, they’re like, can you please get this person out here. So I’m trying to pull this wonderfully nice lady who had had not been able to shower in quite some time, down the hall, and she pulled away from me and stop in the next office and try to sell stuff like you can’t stay, you’ve got to get out. So the next evolution was making sure that our badging system did not auto-unlock the doors and said, we had a little press for service button so that we could capture all that. So yeah, we evolved quickly.

Matt Yokovit: I mean, you’d have to, you’d have to, and over the time, right you’ve evolved, that only your security, but you know, you’ve kind of moved into different areas. So back then it was the early days of a little bit of the Wild Wild West. And so you’ve, you’ve matured as a, as a leader, and I think all of our security standards have matured, so hopefully we don’t see that type of thing happening too often at many companies. But coming in over those years those early days, when did you start to really focus on building your applications using the open-source stack, or using components.

Steve Hoffman: I probably was in about 2010, I had, I joined a company, that, that they did email distribution was what they did, so they built sort of email distribution software. And I had already been a big Tinker a lot of programming on PHP. And so new elements of it, but like, that company showed me bringing all these different pieces together, how much it accelerates the development process I think I’ve been in a couple organizations where it was sort of we pride ourselves on doing it all ourselves, and you just, it’s so inefficient, right, you just don’t have you don’t number one, you don’t have access to the things that you want to change, you can’t make them better, because you’re sort of getting an off the shelf solution, I spent quite a bit of time working for the state of North Carolina, and almost all of our solutions were, this is what you get, right, we’re going to give you the binary CD, you’re going to load it, you’re going to run it, you’re gonna agree to the terms of service, and that’s it, and tweaks and tunes just were not on the menu. So I think, at that company, I started seeing the real value of not just using open-source software for the smaller things that needed to accomplish, but really the big things that you needed to accomplish, and how you could take three or four different products, put them together and get this amazing result. You know, and again I think it also showed me that it comes a with an element of responsibility, right? You can, you can consume all your life, there’s no requirement to give back. But I think for me, there was always a reward in monitoring the forums and someone asking a question, I remember that one was about integrating with multiple active directories and it was just like, holy cow like I had to do this for my company. We got bought by another company, merged our two systems together, but we still have lots of fragments. So we needed both employees from both entities to be able to get into the services and so I just modified the code and it just didn’t dawn on me. That was like The aha moment was, okay, here you go. Like I’ll submit a PR, we didn’t call them PRs at the time, but I’ll submit a PR to get this out there. And lo and behold, it gets picked up. And now it’s sort of the I think they’ve since rewritten it, but it was it became a mainstream option. And this was a WordPress plugin for integrating with Active Directory at the time. So it was really cool to see how that worked. And now, at a place like Percona, the entire business is built around that model, right? Not just consuming, but consuming, sharing freely giving back taking input, working with people, enabling them to do things themselves and sort of without, without your hand open, like okay, I helped you What do you know, what are you gonna do? Without that constraint, it actually makes it free or more, more enjoyable to do it. So

Matt Yokovit: Now, so you mentioned that early time you Percona, and giving freely now, your first kind of jump into Percona was, here’s PMM make it better, right, you’re like, like, your priority is PMS. So Percona, monitoring, management, and Percona monitoring, and management has a pretty lofty vision in the future. And maybe tell us a little bit about where we view PMM evolving to, and then maybe talk to us a little bit about some of those challenges in getting us from where we are today to that?

Steve Hoffman: Well, I mean, if I start two years ago, when I was interested, the PMM product. You know, PNM stands for Percona monitoring and management. But there was zero management. So it was just Percona monitoring like you couldn’t actually do anything to the systems that you monitored, you could, you could only learn that there was something not right. And so one of our biggest transitions is we’re really emphasizing putting that second m into the product line so that we can do management. And we started with an external integration with an alerting system so that you could take a more proactive measure. Now we’ve built that into the tool. And we’ve now actually just released a technical preview of backup management where we can take partner with some of the existing Percona suite, Percona Xtrabackup, and Percona Backup for Mongo. And you can now kick off backups right there, and PMM, which I think is a huge component of systems management. You know, backups are one of those unsexy things that no one gets excited about until they really, really need it, and find out that they didn’t have a great backup strategy to begin with. So that’s part of our vision is number one, adding that second M. And then the other side of it is sort of helping move the market in a new direction a big piece for the PMM product is its Database as a Service offering and, and so we’re really excited to be building that because I think that’s a big enabler for all the people who don’t care about what it takes to get a good setup of a database running. They just, my application needs a port to connect to I don’t care if it’s MySQL, Postgres, or Mongo, I just, I just want to connect my app to the database and do my thing. Please shortcut that. And so you know, Amazon, I think is the gold standard for what they’ve done with RDS. The problem is, there’s, there are quite a few trade-offs with that. And the big trade-offs are, Amazon maintains control of the infrastructure. And as a, as a systems person from way back, you know, there were four vital statistics that I needed to help me understand really where the root of the problem is. So if I can’t see CPU, memory, disk IO, and network load those without those things, I really can’t make a full determination of where do I focus my energy and fixing and an example I had given to someone the other day was as simple as CPU can be 100% pegged. But the solution is different if 100% of that is in user space, or if it’s in system space, or if it’s even in IO wait. So you know, those things have correlations. And they sort of determining where you go next in your investigative path to resolve problems. SoPMM Database as a Service offering actually enables customers to have to maintain that control, right. So you can actually use the limited unlimited scalability of something like Amazon or Google to load your databases on there and sort of extract that away. Your end-user just, I just want my SQL, four nodes, this much memory this much CPU, give me a connection string, get out of my hair, and you know, I think we’re helping bring that and then on the flip side when something goes wrong, your system people, your DevOps folks, your database, people have the ability to that raw access that raw data that will help them pinpoint the problem and go through the refining In the process of either correcting it and software correcting and hardware configuration.

Matt Yokovit: Yeah, I mean, I think that the ultimate vision of being able to enable people is important, right? I mean, PMM already has quite a bit of functionality that enables people to find those problem areas or those issues. So it’s really taking it to the next step. If

Steve Hoffman:
That’s it. That’s it. And we’re not stopping there. Right? I mean, I think I’m talking more about sort of what’s in flight right now. But at the end of the day, I mean, we if Percona has reputation is around sort of our credibility in the market space, telling people well, what is good practice? What is good decision-making? We’re trying to speed that delivery up through things like but I wouldn’t call it artificial intelligence, but programmed intelligence advisors that will be constantly monitoring your system and looking for little optimizations, could you change the configuration, could you maybe modify this query some, is there a better system profile that would give you better performance and throughput, like just little things like that, that we can, that we can be always monitoring for, and with a cloud delivery model, as new ones come out, no action required on your part, just know, yesterday, we didn’t know we had this potential lurking problem. And today, there’s a suggestion that says with just a couple more gigabytes of memory allocated, to your database, which your system already has, by the way, you should be able to get additional performance overhead on your, on your database queries. So things like that.

Matt Yokovit: Okay. And when you are talking to one of the things I know a lot of people want contributions, they want to help with different pieces of software the open-source space is very vibrant and very active. But a lot of people have fear, that they’re not good enough to contribute, or they’re not good enough to work at a company whether it’s us or another open-source company, that they feel a little shy to go out there and help people or to contribute or work at that company, I know that there are some of the smartest people I know who have worked at for Percona, they’ve said hey, for I waited three years before I even applied because I just wasn’t good. Right? Or you see, there’s that mentality that that almost imposter syndrome, if you will, for good developers when you are looking at people who are contributing a lot of code who are being able to push a project like this forward? Whether it is internal or external? What sort of key things do those people have like, what are the markers if you will, that you look for that? You say, hey, that person really that they’ve, they’ve got it?

Steve Hoffman:
Yeah, I can I can give credit to someone who taught it to me. So I worked for a CTO named Ralph Kasuba. This is back at the email delivery company. And he had this philosophy on hiring, that you look first at attitude, then an aptitude, and then dead last edibility. And that, what I’ve learned that the part that he left out is that the right attitude that, that I can get things done that there is no insurmountable challenge, and the right aptitude, just being able to piece together, the things that I do know and apply them to a problem that I don’t know, that can actually overcome the vast majority of challenges. And if you focus there, and hire those people, bring them on board, that solves an awful lot. Now, the ability is, is sort of a piece of it but I think high altitude high aptitude does a lot to overcome modest ability. So in many cases, you don’t need to be a super senior, which I think that’s a struggle that every company I’ve ever worked out with is just always we need to hire the most senior of the smarter the smart. But like, let’s be fair, sometimes those people come with baggage in the form of arrogance. And that’s tough, that hurts a team, you’d rather have people with the right attitude and the right attitude, who with a little bit of investment with a little bit of wiggle room and freedom to grow, do amazing things. And even I’ve seen that even here at Percona. This is something that’s stuck with me since 2011, and I have brought it to every company that I’ve had a leadership role at. And I believe almost all of them are still doing it today. And like I said, I just, it’s just such a wonderful way to look at, at sort of people because you give them a lot more benefit of the doubt. And when people feel like they’re being trusted, that whole fear of I’m not good enough, and I’m not smart enough. If you can give them the competence to just take a stab at it and be blown away with what they can produce. And I think that’s what you see a lot countians are right that thought they weren’t smart enough. Most people probably say, Well, if you can’t answer this super complex question on the spot, you’re probably not smart enough. Instead of saying, use whatever resources are at your disposal figure it, figure it out, take a stab, right?

Matt Yokovit: There’s that curiosity. It’s it’s the willingness to try things, but also be humble about,

Steve Hoffman:
Exactly. Oh, absolutely. Absolutely. I think that’s and, and other people want to work with that, right? Like, I, I get my the biggest energy from a good collaborative environment where we can bounce off ideas, and we can say stupid things. And people say, I don’t think you want to chase that one down. And you’re like, yeah, you’re probably right, I just talking out loud where, whereas I’ve been in other environments where it was crickets because nobody wanted to be the one to throw out a suggestion that might be interpreted as dumb, and therefore, no dumb ideas were ever spoken, but also very few good ideas were ever spoken. So there’s a big trade-off there. And I think if you start building that collaboration and that trust, and you have the right, sort of cultural mindset of we’re gonna achieve this together, you get great things from him.

Matt Yokovit: And with that what sort of contributions or what kind of things are you looking for, for people to contribute back to whether it’s PMM or one of the other projects? I mean, as a engineering the head of engineering here, I’m sure that there’s a litany of things that you want, and everybody always assumes it’s going to be like, this complicated thing. But really, what are you looking for the most?

Steve Hoffman: You know, the take will take anything, I think people assume because Percona is a business, right? A money-making business, that, that we’re only interested in things that directly support our roadmap, well, that’s, that’s nice, I’m certainly not going to say no if you want to work on something that we also want to work on. But at the end of the day what we want is the satisfaction of knowing that we’ve built an extensible enough product that can support what you need to accomplish at the same time is what Percona is trying to accomplish. So I’m always curious to see which directions people are taking it, some things belong on our roadmap, and we never thought to put them there, because we didn’t know. In other instances, we’re just we’re not going down that route. But I think we like helping people give them sort of a here, look, this is how we solve it. For this, maybe that’ll give you a head start on that. No, by the way, if you want, throw it out, that’s and you can save somebody else the effort. Again, it’s not the priority for us. But that’s how we’re going to get to that end goal of being able to support multiple configurations and have lots of options is by welcoming it. And it doesn’t just have to be code contribution. I mean, I told you the very biggest problem that I have in joining an open-source company that shipped software was documentation. Even if you find a typo, or a mistake, or you know, we glossed over some steps, that saves other users hours. And we are sometimes blind to that we’ve just I know the steps. Because I’ve done them so many times, I don’t read the instructions anymore. So the fact that one new step popped in there, all I have to do is mentally insert Step seven, after seven, and I’m good. And if it doesn’t make it into the documentation, we don’t know. So contributions can be as simple as, hey, you missed this step in the documentation, or you had a typo, or this was referring to an old version when I hadn’t fixed that for you. And here’sthe link to the PR. Love it. I mean, that to me, that’s being a part of the community, right? There’s, you can contribute in so many different ways be bringing ideas to light, just log them. in JIRA, we would love to see this fixing documentation, contributing code that we’re working on contributing things that only you care about. All of this is all pieces to that bigger puzzle of open-source software.

Matt Yokovit:
I think the other one that people don’t realize is probably the goal, or the oil of the engineering is feedback. simply telling us how you’ve used something, how you perceive it, why it works or doesn’t work, or how it could be better, even if you can’t hold anything, or contribute anything technical, just saying like, this doesn’t make sense. You know, this is how I think you should approach this. I’ve seen other people do it. It’s such valuable feedback.

Steve Hoffman: Oh, totally. And I think what I like is seeing how other people are getting it to we are here we kind of rely on people submitting feedback via JIRA bug reports or improvement requests, feature requests. We have our forums that we we’re you know, we try to keep our teams very active on the forums. We have our discord channel that you know, that we, again, are fairly active from our engineers that we just understanding what people are trying to accomplish helps us hone in on it. I mean, I think at the end of the day, it we think we’re reasonable Accurate what we’re trying to do, but these are just highly educated guesses and getting that feedback about the problems being solved and why they’re important and what you’re really trying to accomplish. If nothing more than that it either gives validation to our direction, or it helps nudge us a few ticks to the left or right so that we can sort of readjust focus and make sure that you know, it’s not about being Steve’s baby or even Peter’s baby in terms of a product that we deliver. It does satisfy and meet the needs of the people on the other end of the keyboard. That’s who we build the software for. Doesn’t matter if it’s a database server, backup software operator monitoring tool, like, gotta meet your needs. Otherwise, anything we do is just pump our own chest out. You got to get enjoyment out and you got to get useful news out of it.

Matt Yokovit:
Alright, Steve, thank you for stopping by chatting with us a little bit telling us a little bit about your background and about where PMM is headed, and how we can all help.

Steve Hoffman:
My pleasure. Thank you for having me. This was far less painful than I thought it would be.

Matt Yokovit: Always less painful than you thought it would be. In fact, the crowd loves it.

There you go. Wow, what a great episode that was, we really appreciate you coming and checking it out. We hope that you love open source as much as we do. If you like this video, go ahead and subscribe to us on the YouTube channel. Follow us on Facebook, Twitter, Instagram, and LinkedIn. And of course, tune into next week’s episode. We really appreciate you coming and talking open-source with us.

Comments

Did you like this post? Why not read more?

Percona Live 2023 - CFP
✎ Edit this page on GitHub